Chapter 2: Researching sex and gender
How researchers interested in exploring potential differences and/or similarities between men and women conduct their studies is no different than any other field of psychology. As such, in this chapter we will discuss common approaches to the study of sex and gender as well as explore what ethical guidelines research participants must follow when gathering data.
Before any type of phenomenon can be studied, it must first be observed. For example, if a researcher is interested in determining if boys or girls have a stronger or weaker preference for specific activities on the playground, then it would make sense for that researcher to observe both boys and girls engaging in different types of play in their natural environment. Through careful observation, the researcher can note how frequently boys and girls engage in various types of behavior. When a researcher makes observations of individuals engaging in their natural environment, this is called a naturalistic observation. The idea behind naturalistic observation is to be able to see how individuals behave in their natural settings.
The type of data generated from a naturalistic observation is a form of descriptive data. Descriptive data refers simply to information that describes a set of data. It gives you an overview of what the data look like, what the most common values are, and how spread out the data are. It can be shown in simple charts or graphs or through statistics such as the average value. The goal of descriptive data is to give you a general understanding of the data, not to make conclusions or predictions about the data.
More specifically, descriptive data are a type of statistical data that summarize and describe the main features of a dataset. They describe the central tendency, dispersion, and distribution of the data and are often presented in the form of tables, graphs, or summary statistics like mean, median, mode, and standard deviation. Descriptive statistics help to provide a general understanding of the data but do not make inferences about the relationships between variables or make predictions about future observations.
With that qualification aside, the main benefit of naturalistic observations is that it allows researchers to observe the behavior of individuals without the constraints of social pressures which may influence how they act in certain settings. People are often influenced by the social desirability bias which causes them to alter how they behave or what they say to conform to societal norms and expectations. If a boy is interested in the color pink, he may not admit to this preference when in the presence of other boys if he felt that he would be negatively judged for doing so. By removing such constraints, it allows researchers to better grasp the true feelings and motivations of those individuals under observation.
An example of using naturalistic observation to study gender differences is observing and recording behaviors of children at a playground. Researchers could observe the type of games played and how they interact with others to see if there are gender-based differences in social behavior. In the context of this naturalistic observation study, researchers would visit a playground and observe children at play. They would take note of various aspects of the children's behavior, such as the type of games they play, the way they interact with others, and any differences in communication styles. They would also note the gender of the children being observed. After collecting data, the researchers could analyze the information to see if there are any patterns or significant differences between boys and girls. For example, they may find that boys tend to play in larger, more competitive groups, while girls tend to play in smaller, more cooperative groups. This type of research provides a naturalistic setting for observing behavior and can be useful in understanding gender differences in social behavior.
Another interesting example of a naturalistic observation is the research conducted by Dame Jane Goodall (1934 --), an English primatologist interested in learning about the social behaviors of chimpanzees. Up until the late 1950s, most of what was known about the social lives of chimpanzees was based on observations made of chimpanzees in captivity, such as in zoos or research facilities. Arguably, such observations were influenced by the artificial environments in which they were placed. As such, Dame Goodall, starting in 1960, sought out to observe chimpanzees in their natural environment. For six decades, she made meticulous observations of chimpanzees in the Gombe Stream National Park in Tanzania. Her observations resulted in the fascinating discovery that chimpanzees demonstrate very human-like socialization behaviors, an idea which was considered unlikely prior to her research.
"An Adult and Infant Chimpanzee" by Timon Cornelisson is in the Public Domain, CC0
Although the concept of naturalistic observation is quite straightforward, its practical application can be difficult to accomplish. Again, the reason why a researcher would be interested in conducting one is to prevent social desirability biases from influencing an individual’s behavior. However, simply by going to those natural environments in which targeted behaviors occur, researchers alter the true natural state of those environments. How children interact on a playground when they are not being observed may be different from how they interact when they realize they are being observed. In Dame Goodall’s case, instead of observing how chimpanzees interact when they are free of human influence, she became fully immersed in their environment and was directly involved in interacting with them. At best, we can use those findings to suggest how chimpanzees interact with humans in their natural environment compared to how chimpanzees interact with humans in artificial environments. Still, when little is known about a particular phenomenon, a naturalistic observation is a good place for a researcher to start.
In addition to naturalistic observations, researchers can also gather data simply by asking individuals about their behaviors via surveys. A survey is simply any manner in which a researcher gets individuals to respond to questions. A common type of survey is a questionnaire, which involves a list of questions pertaining to various topics. Questionnaires can be easily distributed to multiple individuals and allows researchers to gather a lot of data in a short amount of time. In addition to questions regarding behavior, individuals can also be asked about their thoughts, feelings, attitudes, etc., in an attempt to better understand the phenomenon of interest. For example, a gender researcher interested in comparing attitudes towards seeking out mental health care between men and women could ask not only how many times they have already sought such treatment but also questions regarding stereotypical gender roles and the extent to which they are influenced by common social and cultural norms.
"Focused woman writing in clipboard while hiring candidate" by Sora Shimazaki is in the Public Domain, CC0
An example of conducting a survey to study gender differences is creating and administering a questionnaire to assess attitudes towards gender roles. The survey could include questions about traditional gender roles, such as who should be the primary caregiver in a household, who should work outside the home, and what types of careers are suitable for each gender. The survey participants would be asked to respond based on their own beliefs and experiences, and their responses would be analyzed to see if there are any significant differences between the attitudes of men and women. This type of research allows researchers to gather data from a large sample and can provide insights into gender differences in attitudes and beliefs.
There are generally two types of questions that are included in surveys: closed-ended and open-ended questions. Closed-ended questions (also referred to as ‘close-ended questions) have a predetermined set of possible answers for the individual to choose from. A common example of a closed-ended question is “What is your sex?”, with possible responses including ‘male’, ‘female’, or ‘other’. Personality tests often ask individuals the degree to which statements refer to them on a numbered scale, such as 1 = not at all like me to 7 = exactly like me, with appropriate explanations for the mid-range responses. Closed-ended questions allow researchers to gather specific data with little to no ambiguity as to the responses. Naturally, the quality and accuracy of the data collected is determined by the thoroughness and appropriateness of the provided responses. It is possible that none of the available responses perfectly matches an individual’s way of thinking, so careful consideration should be taken when deciding which responses to provide.
Open-ended questions do not have a predetermined set of responses for individuals to choose from. Instead, as the name suggests, individuals are provided with a prompt to which they can answer however they like. Open-ended questions can be very specific (“What is your occupation?”) or broad (“How does society treat men and women differently?”). These types of questions address the above concern regarding the creation of appropriate response categories as participants determine how to respond for themselves. However, it does require the researcher to code such responses so that general comparisons can be made amongst respondents. Although not exclusively, closed-ended questions lend themselves easily to quantitative analysis (using statistics to draw conclusions) whereas open-ended questions lend themselves more easily to qualitative research (identifying patterns in written responses).
An alternative format to the survey is a focus group, in which researchers gather together a small number of individuals to ask them questions directly. Such focus groups typically follow a pre-determined set of questions aimed towards further elaborating on an area of interest. For example, a researcher interested in exploring potential wage differences earned by men and women performing the same type of jobs could directly ask group participants about their professional experiences, what training or support they received as part of their employment, what their workplace culture is like, etc. An advantage of a focus group is that researchers can ask participants to further elaborate of their responses, which typically is not possible for surveys.
An example of using focus groups to study gender differences is conducting moderated discussions with groups of people to gather qualitative data about gender-related topics. The focus groups could be composed of people from different age groups, cultural backgrounds, and gender identities. The moderator would guide the discussion by asking open-ended questions about topics such as gender stereotypes, experiences with gender bias, and attitudes towards gender equality. The discussions would be recorded and later analyzed for patterns and themes related to gender differences. This type of research provides a forum for participants to share their personal experiences and perspectives and can offer rich and in-depth insights into gender-related attitudes and experiences.
"Photo of Man Sitting in Front of People" by fauxels is in the Public Domain, CC0
All these methods of asking questions of individuals result in the collection of descriptive data. As the name suggests, descriptive data allow a researcher to describe the nature of the information collected. After asking 100 individuals to complete a survey, the researcher can tabulate how many men versus women participated, their average age, what percentage of respondents identify as which race or ethnicity, etc., based on what questions were asked. These data allow researchers to quantify massive amounts of information and distill them into simpler concepts. Asking the same questions to different people over a period of time could help determine shifts in common thinking amongst a population, such as shifts in attitudes between men and women towards the division of household chores.
The use of descriptive data has several advantages. Descriptive data can reveal patterns and trends in a dataset, allowing researchers to gain a better understanding of the data. The data are usually presented in a simple and easy-to-understand format, such as tables or charts, making it accessible to a wide audience. It is also cost effective as collecting and analyzing descriptive data is often less expensive than other research methods, making it a cost-effective option for research projects. Also, descriptive data can be collected from large sample sizes, providing a more representative view of the population being studied while also providing a starting point for further analysis and can provide initial insights that can guide future research.
A drawback to descriptive data, however, is that the researcher is limited to reporting just the information that was gathered. It would be inappropriate to infer how another person might have responded unless such information was collected. For example, a developmental researcher could collect information from seven- and nine-year-olds regarding their moral decision-making processes but would not be able to use this information to predict how eight-year-olds might respond (such as by averaging their responses). As such, descriptive data are extremely important and can provide a wealth of information, but such information is also narrow in scope.
Often, a researcher is interested in going beyond the data collected in order to extrapolate from the information obtained. For this to occur, all that is necessary is for the researcher to gather at least two forms of descriptive data. By doing so, it is possible to conduct a statistical technique known as a correlation. A correlation identifies how two pieces of information are related to one another. These pieces of information are referred to as variables. In other words, if a researcher collected information on people’s ages, then age would be a variable.
Once a researcher has two variables, the correlational analysis will indicate the nature of the relationship between them. This analysis results in an r value (technically, a Pearson product-moment correlation coefficient). These values range from -1.00 to +1.00 and indicate both the strength and direction of the relationship between the variables. If the r value is positive, i.e., above 0.00, then this indicates that as the value of one of the variables increases, so does the value of the other variable. This is referred to as a positive relationship.
Let’s consider a hallmark example of a correlational analysis involving two variables: high school GPA and college GPA. In this scenario, a researcher measures a high school senior’s GPA upon graduation, and then later measures the same student’s GPA upon matriculation from college. The researcher now has two measures of descriptive data. After running a correlational analysis, let’s say the result is an r value of +.65. This would indicate that students with higher GPAs in high school will also have higher GPAs in college. (The same is also true of standardized academic tests such the SAT or ACT; higher scores on either are associated with higher GPAs in college, which is why some schools are interested in having students submit this information).
An example of using correlational research to study gender differences is examining the relationship between gender and career choice. Researchers would collect data on both the gender of the participants and their career choices, and then use statistical techniques to assess the strength and direction of the relationship between the two variables. For example, the researchers may find that women are more likely to choose careers in education or healthcare, while men are more likely to choose careers in engineering or technology. This type of research can provide insights into potential gender differences in career choice, but it cannot establish cause-and-effect relationships. Correlational research can be useful for exploring relationships between variables and identifying potential areas for further investigation.
As such, the advantage of a correlational analysis is that it produces predictive data which allows the researcher to go beyond the data collected. This is because the correlational analysis results in a best-fit line, an indication of the overall relationship between two sets of values. Knowing this relationship does allow a researcher to take just one piece of data (high school GPA) and predict the value of a second piece of data (college GPA). Even if the original dataset did not include a particular high school GPA, the research can still infer what college GPA should be associated with it.
On the other hand, if the r value is negative, i.e., below 0.00, then this indicates that as the value of one of the variables increases, the value of the other variable decreases. This is referred to as a negative relationship. Consider the relationship between smoking cigarettes and physical health: the more an individual smokes cigarettes, the lower his or her physical health. Or, in terms of gender, the more strongly an individual condones traditional gender stereotypes, the less likely that same individual would believe in egalitarian relationships.
It is also possible that there is no correlation between two variables, i.e., an r near or at 0.00. This would indicate that knowing the value of one variable would not allow you to predict the value of a second variable. For example, knowing whether or not a person wears reading glasses would not allow you to predict what type of vehicle that person drives.
In addition to whether the correlation is positive or negative, as indicated by the r value being either positive or negative, the number value itself indicates the strength of the relationship. The higher the value (regardless of direction), the stronger the relationship between the two variables. The closer the value is to zero, the weaker the relationship between the two variables. For example, an r value of .81 is stronger than an r value of .54. Conversely, an r value of -.67 is stronger than an r value of .33. In other words, the direction of the relationship (as indicated by being either positive or negative) should be considered separately from the r value itself (the number which ranges from 0.00 to +/- 1.00).
As such, predictive data are more advantageous than descriptive data in that they allow researchers to go beyond the data provided and infer relationships even when specific data are not collected. However, there are some limitations to predictive data as well. Although correlations allow us to better understand how two variables are related, it does not indicate how or why those variables are related. As commonly taught in statistics classes, “correlation does not imply causation”. In other words, knowing that exercise and mood are positively correlated does not allow us to suggest that exercising causes us to be in a better mood. It is just as likely that being in a good mood in the first place motivates us to engage in higher levels of physical activity. Although it can be very tempting to suggest that one variable causes another to change in a predictable manner, there are many reasons why this must be avoided. Just because a meteorologist can predict whether it will rain or not does not imply that the meteorologist causes the rain to fall.
In order to understand how variables are related to one another, such as what causes one variable to affect another, it is necessary to conduct experiments. Unlike surveys where variables are only measured, experiments involve the manipulation of one variable under controlled settings to ascertain if any changes occur to a second variable. It is this manipulation of variables that allows a researcher to determine causation.
To better describe how experiments are conducted, a distinction has to be made regarding which variables are included in a research study. In general, there are two main variables in a research study: an independent and a dependent variable. The independent variable is the variable that is manipulated, or made to be different somehow, and is theorized to be responsible for causing changes to another variable. Manipulation can be as simple as the presence or absence of the variable, or varying the degree to which the variable is present. The dependent variable is the variable that is expected to change as a result of the manipulation of the independent variable. In other words, this variable changes depending on what happens to the independent variable.
In the simplest case, a researcher can manipulate an independent variable by altering its presence or absence in a research study. For example, in pharmaceutical research, the effectiveness of a new potential drug is examined by offering the actual medication to one group of participants while offering a placebo (a similarly shaped pill but without the medication) to another group of participants. In this situation, the presence or absence of the medication is the independent variable and the effect on one’s health is the dependent variable. This type of study would then allow the researcher to determine if the medication causes a change in a person’s health. The group receiving the medication should experience positive changes to their health whereas the group receiving the placebo should not experience any change to their health. If there is no change in physical health in both groups, then the medication would not be considered to be effective. Only in those situations where the manipulation leads to a change in the dependent variable can causation be determined.
For phenomena that have already been investigated, the presence/absence of the independent variable can be replaced with varying the degree to which the independent variable is present. If an experimental drug has been demonstrated to be effective compared against a group of individuals who did not receive it, a follow-up study could then be conducted to determine what amount of the medication is most effective. Researchers could then give prescribe varying amounts of the medication to various groups to determine what dosage has the greatest positive effect on physical health. This is the process by which appropriate dosages of new medications are determined.
Regardless which approach is taken, researchers must carefully consider how confounds can influence the results of their studies. A confound is any variable that interferes with the ability to determine causation. Most research studies are conducted in tightly controlled settings. This is to ensure that the only difference between two research groups is the change in the independent variable. If there are additional differences between groups that the researcher did not control for, then the ability of the researcher to determine causality is significantly lessened. This is because any changes between research groups could be attributed either to the manipulation of the independent variable or due to the presence of the confound.
There are several types of confounds that must be controlled for in an experiment being conducted to determine causation. For example, a history confound involves changes in the environment that occur simultaneously along with the manipulation of the independent variable, and is particularly problematic when a lot of time passes between the manipulation of the independent variable and the measurement of the dependent variable. Imagine a researcher is interested in determining if changes in one’s diet affects one’s overall psychological well-being. As part of this investigation, the researcher recruits a group of participants with similar profiles and puts them on a calorie-restricted diet for six weeks, with the intent of measuring changes in their psychological well-being from the beginning to the end of the study. During the six weeks that the study is taking place, however, it is possible that outside events could also affect participants’ well-being. The weather could progress from dreary, rainy days to warm, sunny days over the six weeks. Participants could join new social groups that increase their amount of friendly interactions during their free time. Perhaps participants adopt a rescue puppy. Any of these events which occur simultaneously along with the introduction of the independent variable would lessen the ability of the researcher to determine causation as any change in psychological well-being at the conclusion of the study could be due to either the change in diet or any of these other factors. In addition to recruiting a group of participants who have their diet changed, the researcher would also need to recruit a separate group of participants who do not change their diet over the same period of time. As both groups of participants would be subject to the same influence of these outside factors, causality could only be inferred if a meaningful change in psychological well-being occurs in the group receiving the experimental manipulation, i.e., the change in diet.
Another confound to consider is maturation, which refers to internal changes in participants over time. Consider an academic intervention aimed towards increasing a primary school’s scholastic abilities. A researcher identifies a group of students with low academic performance and assigns them to weekly tutoring sessions over the course of a semester. At the end of the study, the researcher then compares the academic performance of the students from the beginning of the study to the end of the study. Even if the researcher finds that student’s overall academic abilities have improved, it is still not possible to determine if the tutoring sessions caused this change. This is due to the fact that, over the course of a semester, these students have matured in terms of their experiences and are simply no longer the same individuals that they were prior to the start of the study. It is just as possible that their overall life experiences are responsible for the change in their academic performance as it is that the weekly tutoring sessions are responsible. Similar to how to control for a history confound, the researcher would also need to recruit a similar group of participants who do not receive the weekly tutoring sessions. Only if the group who receives the tutoring sessions demonstrate a greater change in academic performance compared to those who don’t would the researcher then be able to establish that the tutoring sessions cause a change in the academic performance of the students.
A third confound to be aware of is instrumentation, which refers to the manner in which the dependent variable is being measured. Whereas history and maturation confounds are more closely associated with the manipulation of the independent variable, an instrumentation confound is directly related to the measurement of the dependent variable. Instrumentation confounds can occur in one of two ways. The most common example is when there are changes made in how the dependent variable is measured from the start to the conclusion of the study. If a researcher studying happiness first defines it as the number of times that a person smiles during a conversation, but then later defines it as how well they score on a happiness questionnaire, then it would be impossible to determine if any type of manipulation of an independent variable (such as watching a comedy film or not) could be responsible for causing a change in happiness. Any difference in happiness in the study could simply be attributed to differences in how happiness was measured and not as a result of the introduction of a manipulated independent variable. The obvious way to control for this type of confound is to ensure that the measurement of the dependent variable remains constant throughout the study.
Another type of instrumentation confound can occur when there is more than one person measuring the dependent variable. Suppose a researcher is interested in evaluating how sociable children are during recess on the playground. The researcher recruits two assistants to observe the behavior of a group of children during a fifteen-minute recess period. One assistant records the number of times that a child engages in conversation with others; the other assistant records how many children are grouped together at one time. Any difference in the amount of sociability that the assistants come up with could simply be due to differences in how said sociability was measured. To control for this type of instrumentation confound, the researcher would need to ensure that both assistants are measuring sociability in the same way.
A fourth confound can occur when there is a statistical regression to the mean, in which extreme scores on a measure become less extreme when measured again. Let’s say that you were interested in determining the effectiveness of a new type of bowling technique. You want to investigate how well this technique improves the performance of both expert and amateur bowlers. As such, you recruit a group of participants and have them each bowl a game. You then divide the high scorers from the low scorers. The problem in this scenario is that you do not know how well their performance in that one game accurately matches their general level of skill. It is possible that how well they perform in that game is fairly indicative of their general abilities. However, it is also possible that their performance in that one game represents either their absolute best or worst score ever in their lives. If the latter scenario is true, then it is extremely unlikely that if they were to immediately bowl another game that their score would the same. If their first game represented their all-time best score, then most likely their second game would result in a lower score. On the other hand, if their first game represented their all-time worse score, then most likely their second game would result in a higher score. In both scenarios, the extreme score from the first game would become less extreme in the second game. Regardless, the researcher would not know whether the first score is truly representative of their bowling ability or not, which may lead to inaccurately reporting their bowling status. Instead of solely relying on their performance in just one game, the researcher would benefit from having each participant bowl at least two games to determine if there is any extreme deviation in their scores. Only those participants who bowl somewhat consistent scores would then be included in the study, whereas participants with wildly different scores would be excluded.
"Assorted Bowling Ball Lot" by Matthias Zomer is in the Public Domain, CC0
A mortality confound occurs when an unequal number of participants drop out of a study. Particularly for studies that extend over a significant period of time, it is expected that some participants who begin a study will not fully participate through the end for various reasons, e.g., moving away, losing interest, changes in eligibility. Seasoned researchers already know that it is generally necessary to include a minimum of two groups in any experiment in order to control for history and maturation confounds. Let’s say a researcher is interested in determining the effectiveness of a two-hour study group each Friday at a local high school. The researcher randomly divides 100 participants into two groups of 50, with one group participating in a two-hour study group while the other group sits quietly in a room for two hours. The study is intended to last for six weeks. Prior to the study, all students are tested and it is determined that they have, on average, equal academic ability. After six weeks, the researcher again tests the students and finds that the group receiving the study sessions performed higher than the group who did not receive the study sessions. Although it may be possible that the study sessions caused a positive change in academic performance, the researcher should pay careful attention to how many students remained in the study over the course of six weeks. The students in the study session group may have realized the potential benefits of their participation and decided to stay for the duration of the study. The students in the non-study session, however, may have decided that they had better things to do on a Friday afternoon and thus decided to drop out of the study at some point during the six-week period. If there is greater participant loss in one group than there is another, then statistically the two groups are no longer considered equivalent. Any changes in the dependent variable that arise may be due to the introduction of the independent variable or simply due to the difference in the resultant composition of the two groups.
The generation of explanatory data is essential to understanding how variables are related to one another. The ability to explain how the manipulation of one variable can cause a change in a second variable is the cornerstone of empirical research.
Thus far we have discussed how to generate three types of data, including descriptive, predictive, and explanatory data. Each type of data plays an important part in better understanding the relationship between and among variables. However, it is not always necessary to generate data to better understand a phenomenon. Indeed, most phenomena of interest have already been studied to some degree or another by other researchers in the field. As such, another tool in research is to rely on the results of studies conducted by other researchers.
Archival data simply refers to the existing research literature and can take any number of forms. Old newspaper articles saved on microfiche, census records, books, and published journal articles are just some examples of archival data. Indeed, prior to conducting a study, researchers must carefully read through the already existing literature to ensure that their efforts are not being wasted simply by replicating a study that has already been conducted. Archival data also provide justification for why a new study is being conducted. They can be used to identify where gaps in knowledge exist and how a new investigation can help to fill in those gaps.
When using archival data to study gender differences, researchers will examine historical records, documents, or data that have already been collected and stored. These may include census records, government reports, newspaper articles, historical texts, and other forms of written or recorded data.
By analyzing these historical sources, researchers can gain insights into gender differences across a range of topics and over a longer period of time. For example, by examining census records, researchers can track changes in the education, employment, and income levels of men and women over the course of several decades. By analyzing newspaper articles or historical texts, researchers can understand how gender roles and expectations have changed over time, as well as how women's experiences and contributions have been portrayed in different periods.
Archival data can provide a broad historical perspective and can offer valuable insights into the past. Additionally, because the data has already been collected, researchers can focus their efforts on analysis and interpretation, rather than data collection, making this type of research cost-effective and time-efficient. However, it is important to keep in mind that archival data may not always be complete or accurate, and that the interpretation of the data may be influenced by the researcher's own biases and perspectives.
The advantage of using archival data is that there is a wealth of information that is readily available to a researcher without having to recruit participants, collect data, run statistical analyses, etc. It is quite possible that other researchers have already investigated particular phenomena, so it is important to first understand what the results of previous efforts have been before engaging in further studies. It would not behoove a researcher to simply replicate the efforts of others who have already come before them. Instead, it is much more preferable to continue to add to the diversity of knowledge in an area of study such that further insights can be ascertained.
The disadvantage of archival data is that previous researchers may not have collected the type of information that you are interested in. For example, whereas transgender rights are discussed quite frequently in today’s news, fifty years ago it was not as common. A researcher interested in understanding how societal attitudes towards transgender rights have developed over time would have great difficulty relying on archival data to address this issue.
In addition to being mindful of the types of data gathered for a study, researchers must also ensure that the rights of participants who are supplying such data are being protected. To ensure that these rights are properly upheld, psychological researchers must follow standard ethical guidelines which have been updated through the years.
In general, the benefits of any research study must outweigh any possible risks to the participants. These risks include a loss of confidentiality, physical or mental harm, embarrassment, potential legal liabilities, and a host of others. Participants must be fully informed of the potential risks involved with their participation in a research study prior to giving informed consent, an explicit acknowledgement of the details of their participation. This informed consent must provide information regarding what is expected of them as a participant, what alternatives (if any) there are to participating, if there is to be any sort of reward or compensation for their participation, any potential risks, if any support services are available, as well as the contact details of the principle investigator of the research study. Federal law dictates that participants must be between 18-90 years of age in order to provide consent.
The researcher is not the one who determines whether the cost/benefit of the study justifies whether it is conducted or not. Instead, most institutions have either an ethics panel or an IRB (Institutional Review Board) who objectively and independently makes this determination. Such bodies ensure that the rights of participants are being upheld, that all local, state, and federal laws are being followed, and that the potential results of the study justify it being conducted.
Ethical principles play an important role in ensuring that gender research is conducted in a responsible and respectful manner. One such principle is to ensure respect for persons. Participants in gender research must be treated with respect and dignity, and their rights and autonomy must be respected. This includes obtaining informed consent from participants, ensuring confidentiality, and avoiding any harm or exploitation.
Research must be designed to maximize the potential benefits and minimize the potential harms to participants. This principle is known as beneficence. Researchers must be transparent about their methods and should not engage in research that is likely to cause harm to participants.
Research must also not discriminate on the basis of gender, race, ethnicity, religion, sexual orientation, or any other characteristic. Researchers must avoid engaging in research that reinforces harmful gender stereotypes or reinforces gender-based discrimination.
Researchers must maintain the trust of participants and adhere to ethical standards, avoiding any deception or manipulation. Researchers must also report their findings honestly and accurately, avoiding any falsification or fabrication of data. Researchers have a responsibility to consider the broader impact of their research, and to act in the best interests of society. They must also ensure that their research is conducted in an ethical manner and take steps to address any negative consequences that may arise from their research.
In conclusion, gender research must be conducted in a responsible and ethical manner, upholding these general ethical principles. This helps to ensure that gender research is conducted in a way that respects participants, maximizes benefits, minimizes harm, and does not reinforce harmful gender stereotypes or reinforce gender-based discrimination.
Imagine a child development researcher interested in learning about any potential gender differences regarding childhood play. Here’s an example of how a researcher approaching this topic from a naturalistic observation perspective would differ from another researcher approaching the same topic from an experimental perspective:
A gender researcher using naturalistic observation to study the differences between how boys and girls play could go to a park or playground and observe children as they play without interfering. They would take notes on the types of play that each gender engages in, such as the toys they choose, the way they interact with others, and the physical activities they participate in. They would also look for patterns and similarities in the behavior of each gender and compare their observations with existing theories on gender and play. The researcher would then analyze their data to identify any significant differences or trends in the way that boys and girls play.
The researcher might also consider various factors that could influence the children's play, such as age, culture, socio-economic status, and family background. They would take these factors into account in their observations to ensure that any differences they identify are truly related to gender and not other factors. The researcher would also aim to observe a diverse sample of children, including those who may challenge traditional gender stereotypes.
In order to ensure the validity of their findings, the researcher would follow scientific protocols, such as using a large sample size and keeping detailed, unbiased notes. They would also cross-check their observations with those of other observers to ensure consistency.
Once the data have been collected and analyzed, the researcher would write up their findings and discuss any implications for our understanding of gender and play. The results of the study could be used to inform early childhood education programs and to guide future research in this area.
On the other hand, a gender researcher interested in studying the differences between how boys and girls play from an experimental perspective might use a controlled laboratory setting and manipulate the conditions under which the children play. For example, the researcher could randomly assign children to different play conditions such as playing with a particular toy or playing with peers of a certain gender.
In this type of study, the researcher would control for extraneous variables and use a randomized design to increase the validity of their findings. They would also measure the behavior of the children using objective measures, such as the amount of time spent playing with each type of toy or the number of physical activities engaged in.
The experimental design allows the researcher to make causal inferences about the relationship between gender and play behavior. By manipulating the conditions under which the children play, the researcher can determine whether any observed differences in play behavior are due to gender or to other factors.
Once the data have been collected, the researcher would analyze the results using statistical methods to determine whether there are significant differences in play behavior between boys and girls. The researcher would then write up their findings and discuss the implications of their study for our understanding of gender and play.
There are advantages and disadvantages associated with either approach towards studying the play behavior of children. In terms of the advantages of using naturalistic observation, this approach provides a more authentic and representative picture of the phenomenon being studied, as children are observed in their natural environment. Researchers can observe behavior over an extended period of time, allowing for a more in-depth and nuanced understanding of the behavior. It also offers a more flexible and spontaneous approach, allowing the researcher to adapt their observations as new and unexpected behaviors emerge.
However, when conducting a naturalistic observation, it is difficult to control extraneous variables which can affect the validity of the findings. Observer bias, even unintentional, may influence the way the data are collected and analyzed. In addition, it may be challenging to observe a large enough and diverse enough sample of children.
In terms of the advantages of using an experimental approach, this allows a researcher better control of extraneous variables, which increases the validity (or accuracy) of the findings. This approach can also be used to establish cause-and-effect relationships between the independent and dependent variables, meaning those variables that the researcher manipulates as well as measures. Experiments also offer a more structured and systematic approach which can lead to more objective and accurate data collection and analysis.
However, experiments may not accurately reflect real-world behavior as children may behave differently in a laboratory setting. In addition, the artificial nature of the laboratory setting may limit the generalizability of the findings to other settings or populations.
Given the advantages and disadvantages of both approaches, using multiple research approaches is important because it provides a more robust and reliable understanding of the phenomenon and helps overcome the limitations of individual approaches. Each research approach has its own strengths and limitations, and by using multiple approaches researchers can triangulate their findings and address the limitations of individual approaches.
For example, if researchers use only naturalistic observation to study a phenomenon, they may miss important variables that are better controlled in an experimental setting. Similarly, if researchers only use an experimental approach, they may not get an accurate picture of real-world behavior. By combining both approaches, researchers can benefit from the strengths of each and overcome their limitations. Additionally, using multiple research approaches can also increase the validity and reliability of the findings. By cross-checking the findings from different approaches, researchers can confirm or refute their conclusions and gain a deeper understanding of the phenomenon being studied.